[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: fsck journal replay times (was Re: Is this list active?)



Theodore Tso wrote:
> 
> On Sat, Jul 14, 2001 at 04:07:34PM -0400, Ed McKenzie wrote:
> > ext2/3 mark the fs as damaged if they encounter anything strange after
> > mounting.  Wouldn't this catch most fs corruption and force fsck as
> > needed, without unnecessary checking on consistent fses?  That's been
> > my experience, anyway -- ext2 has always been the one to catch
> > corruption, and fsck's mount count limit has been somewhat less
> > useful.
> 
> Not all filesystem corruptions will be caught by the filesystem kernel
> code.  For example, if some bits in the block allocation bitmap get
> cleared, blocks in use by one file may be reallocated and reused for
> another file.  Usually, this will never be noticed by the ext2 kernel
> code.
> 
> I've definitely had cases where after the mount count went over the
> threshold, e2fsck has detected multiply claimed blocks.

Now that buffer_head has b_inode, could we not add runtime
checks for this?

We'd need to populate b_inode for reads, but that looks to be
OK.





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]